I'm going to let this one flow and only leave one question NO ONE thinks about:

How does the Physics in a X-over Actually work? How a Capacitor or Coil actually filters.

Answer that tuffluck and your adding makes no sense.

I liked these discussions because it Airs falsehoods and general misconceptions in the Car Audio Industry. Good stuff by everyones accounts. The Last discussion went But some valid points and questions were brought up.

Here's the old thread and let the discussion begin.

https://ls1tech.com/forums/showthrea...6&page=1&pp=20

 

Wow i feel famous, my thread is referenced in another thread!!!

 

lol. Yeah. I like it when someone (you in this case) brings up a Valid question.

I do believe your question was answered, but then the Topic ran off into the Technical aspects of x-overs and their designs. That is what I'm trying to get started here.

I've fallen prey to alot of misconceptions in the past and it took a while to burn out the Bad Habits. Ever Heard someone Say "Cheating the Laws of Physics"? That is a really bad one to hold onto and I actually did when I was in High School.

 

Hi guys, long time no post over here...

If the reactive device is in series with the device being filtered its rising/falling impedance with frequency acts as a voltage divider and thus cuts power delivery to the device based on frequency. If the reactive device is in parallel then its rising/falling impedance acts as a variable shunt current divider to again cut the power delivery to the device in question.

So in short its a variable resistor that automatically changes based on frequency. We take that resistor and use it in series parallel to get our desired response (and end up with phase changes and other fun junk).

Now a question back to you: What is a Linkwitz-Riley 24db/oct filter besides two cascaded 12db/oct Butterworth alignments?

No peeing match here guys - yes the cascade will cause there to be a -6db point at the crossover (which then requires proper phase control to get correlated constructive addition for flat power response) but there's also the idea that the crossover point could be slightly shifted to compensate for that and overlap the rolloffs to provide a -3db point if the driver's output is noncorrelated.

But there is still the main issue in the car to deal with - the acoustics. A couple db in a region where most cars have a boatload of boost isn't likely an issue. How many guys are running EQs to help control their overall tonal balance? Or even better, don't have their bass cranked way up? Its a pretty small margin of error to double up on a crossover most likely intended for some big honking car sub that doesn't play above 100Hz where our ability to discern distortion is incredibly poor.

my 2 cents...

 

good topic.

all that was so over my head its not even funny.



anyway, i use my head unit (alpine 7995) crossover as my main xover, and enable my amps crossovers but put them on the lowest settings (say 40hp for component amp and 250lp for sub amp). i've tried it every way, and i didnt notice any difference other than having to set my midrange a little higher to keep 'em from distorting when i used the amps xover or the head units by itself.

 

I think we need pictures and diagrams, thats what works in all my physics textbooks. I can figure everything out if it has pretty colors and all that ****.


Bring on the diagrams!!!!!!

 

alpine deck crossovers have a relatively slight slope; with my bass amp crossover set to the lowest setting and the HU xover at 80Hz, the bass is ridiculous. put the amp on 80 also, and you actually get clean sound.

 

I notice that too with my setup. If the amp and HU are set somewhere around 78Hz (highpass), there is considerably less bass than say when the amp crossover is set to 100 to 150Hz and the HU crossover is turned off. It seems that one crossover just isnt doing enough below the target frequency.

 

I just hook **** up till it either smokes or sounds good. Either way, I get *some* entertainment out of it.

(kidding)

Truth be told, when I started doing this professionally (over a decade ago), there wasn't much technology in the car audio field. I had been in pro audio prior to that, so much of the theory and principles applied, just on a smaller scale. It's amazing how much there is to know once you really dig in. Today's consumer electronics have made tuning much simpler by integrating the built-in crossovers that you guys mention - back in the day, we hand-built our own passive crossover boards with various discrete components (resistors, capacitors, chokes, etc) to achieve the different slopes we wanted. Fun stuff, but lots of math.

 

An RC circuit.

I looked over all those old posdts of yours. I mean DAMN too many of them

Nailed it dead on the Head. I think some people's brains are going to hurt by that explaination. You have an E.E. or Tech background?

The original pissing match basically started when someone said that it is OK to just add a 12 dB and 18 dB circuit because it would make for a larger roll off. Something like that. Anyways, To add back to back SAME crossovers is a bad idea.

 

Went to school for EE at a state college in ND, pretty sad program (very long story short) but been doing computer programming/support since and an avid audio hobbist for a long time (its more fun than code).

I don't see any issues with adding additional crossovers of any type to modify the signal really, and the amount of change isn't always as labeled (and I mean more than just how poor some of the ***** match the numbers). And to a certain extent, its all the EQ is really doing anyways except it affects a narrower band.

We had an AudioControl DQX in our car and swapped it out for an Alpine H701. Now being digital EQs they have all kinds of 'liberty' in their algorithms but the difference is substantial. At some point I'll hook them up and scope them to find out who is lying more but a 24db/oct slope should be just that and Xdb boost/cut should be too. Yet they are like very different animals in how they sound. Puzzling.

It was so much easier 15 years ago. You got a 2XS and a Punch 45 for the fronts and a Punch 60 on the subs and rocked away! Maybe an Alpine EQ and some PPI amps, but it seemed so simple. Now its like 1/2 the price and 4x the features, most of which sound like junk. Technology!

 

Cool. I can Probably count on insite from ya. I read the Capacitor Arguement And That's another One I'm going to Bring up next. I don't claim to know everything, but I have Been Tinkering With Electronics for about 16 years straight now. My Army Tech school That I finished was a Year Long (11months, 8 hours a day, 5 days a week) I have been istalling Car Stereo Equip for about 14 of that. I spent 3 years in Local 134 (Electrician) and I've been working on Electronics as a Tech for almost 9 years straight.

They Use IC coupler and that is probably where the Change in sound is. Also the Designer follow 2 different Paths of Design: Some say a 24dB/slope should be a Bessel (0.5 AVG) and not the Butterworth (0.707) design. That will change the Sound design and TAilor the Phasing to be different.

Yeah. I kind of agree. I remember when Everything was Easier and Harder all at the Same time. I still have a Punch 45 and Punch 150 sitting in my Garage and I don't want to get rid of them.

Oh. Alpine 7909. RF Power 1000, ****. I should make a "Blast from the Past" thread. Brings back memory.

 

Right. They are Notch filters. That's a good point. The Path of design is: Input signal, Tailor sound and output plus Gain. But it is Modified within a single Chasis and the Sound is Modified accross the Spectrum.

The Only Problem I have is the Combining of Filters at the Same Cutoff points. It is Almost the Same effect in Gain as Recording from Tape to Tape and Adding +3-10 dB worth of Gain and then Doing it Again. It will Most definitly Amplify the Noise and Ripples at the Crossover point.

Also, the Phase Lobing and Time Shift's are MUCH harder to deal with.

The Original Response was to Add to simply Add. For say the Sub range. Now the rest of the System is Modified Separetly from this Part of the Spectrum. If you look at from this point, it then Decides to Separte itself from the Spectrum due to Time Lag and The Lobing Difference.

For the Sub range in a Vehicle that probably won't make a difference, but for ANYTHING above 250-300Hz the Axial Response and the Time Lag are of Critical Possitions.

The Whole Deal with the Linkwirtz-Riley becoming popular is it Addresses the Problems associated with just doing that.

What do you think???

P.S. I think this is going to turn into the 2 of us soon. Oh. I'll try to dig up some info on these points. I maybe wrong, but I'm going to find out one way or the Other.

 

Damn. That might prove my point entirely. Slew Factor is a Major point of Difference between Transistors. Might be the Slew is Slower than the other and that would correspond to the Sound change.

Or it is still along the Lines of The difference in Design Phillosophy. The AVG vs RMS calculations. (Basically if you don't know. (Thrown out to everyone, not Cliff_J) Physics has Transversed into more Calculus Based Equations (AVG) vs Algebraic (RMS) based equations.)

 

Agree completely, maybe even down to 150Hz on really good material with great resolution and up-front stereo subs for the phasing thing. Now where do I find a spot for the HVAC in the back and get enough airflow to heat and cool the car so I can get some nice speakers up front?

As far as the time difference, mmm, I'd guess a little higher but it could be noticable by 300Hz and up if it was really bad. Its difficult to do even 5 msec of delay in the analog domain, I don't think it'll be anything close to that in most designs but there is always the potential to muck things up big time! So if it isn't picked up but a frequency response output then it shouldn't be of importance.

Yes, it may very well be the two of us already.

Regarding caps, I think its really funny to read the old autosound 2000 tech briefs and it discusses how you can hear the difference of adding a high-quality cap when listening to something like l'daddy. And now he offers $5k if you can hear the difference adding a cap (although I believe he restricts that to the new carbon caps) but it would be interesting someday round up an old Alpine 3545 and see if it really does make a difference. Its one of those never ending arguements on carsound's forums like if amplifier's properly matched sound different or if wire sounds different. I've heard the difference before but once I understood the science it made sense why it happened.

I don't know exactly on the sound difference between a DQX and H701. Anyone can get a tech sheet from a manf and A/D and then D/A, but which DSP and what algorithm you use is another matter all to itself. That stuff can cure insomnia and its difficult to make sense of any of it as a hobby.

I thought at first it might be related to the cost savings in the H701 with its use of a single 8-ch D/A and crappy volume chip (it tracks fine L/R but it doesn't respect zero crossover points so it 'studders' the volume sometimes) but a couple guys with the H900 F#1 processor have remarked too that the boost/cut doesn't seem the same as it would on an AC piece either. And the H900 has pretty much all the nice analog circuitry a guy could want with Burr-Brown utilized throughout, it had better for the price. So I'm leaning towards blaming their software in the DSP rather than any analog processing, there are some odd methods they chose to clock the inputs and what not that may cause jitter problems and so on.

I really would like to convert the powersupply to 12VDC and play with a Shure 4800, Rane RPM26, dbx 260, or Behringer DCX2496 - not only can you choose the slope up to 48db/oct but on the Rane and dbx you can choose the alignment you want and do full parametric EQs. I'd love for someone to make a DSP in-a-box kit with some high-level langauge to program it. It'd be awesome to write software to mimic the exact nature of a analog circuit in digital. There's even a processor out there from DEQX that can do 300db/oct filters!! Too bad its $4k!! But the possibilities with digital processing are pretty unreal regardless.
http://www.deqx.com/dsp-product/pdc26.html

 

Damn. I guess we can guess and Therorise all day long. That is why "Tech Labs" Exist.

I love that Rane piece. I remember back in the Late 90's that was the Piece to Own for SQ. Too bad that is WAY beyond most people and that kind of Money is what drove me away from Competing until SPL came around and I jumped on the Ported Bandwagon Early.

Oh well. May a X-over Part III to get more User friendly info out? Help me write it up?

I have been wanting to create an Installer and Advanced Portions within the FAQ. I'm thinking the Install area would be a Good idea, but the Advanced part might be good to do in this way.

Hell. I like this because I haven't had dicussions like this in a While. Really pushes the Grey matter.n

And, Welcome back

 

Dayem Cliff, can you please be a little more specific?

You guys should have pre-posted, FOR ET's only! (Electronics Techs)

I followed about 15% of it, and that is good for me! I agree w/you guys in that analog and passive x-overs were like a 1969 GM and NO electronics(much easier to tune/service).
Today, w/digital and consolidated circuitry performing multiple functions, it is much harder IMO, but I am NO TECH, nor do I claim to be for the record.

You guys seem to know your stuff

 

Yeah. This one Stayed on topic, but went for the Nobel Prize. Way too indepth for most people. I'm a Nerd. lol

I'm finnally going to get the Installer FAQ stuff done. It WONT go indepth like this.

 

i still don't see why there is anything wrong with running an amp crossover and a HU crossover together, guys

 

There isn't anything wrong with it. With one caveat: Unless you want it at a frequency where you could hear it like on spoken voice or piano or what not - lets just say above 150Hz or so. Anything below that is fair game to go ahead and try it to see if it works ok in your setup. Above that and its ancient alchemy (or to use a cliche, throwing mud at a wall) where the results of blind trial and error are more likely to fail than succeed.

Its like asking if there is anything wrong with using regular electrolytic caps (without a mylar bypass cap) and inductive resistors on a passive crossover. Maybe if your tweeter has a lot of output on the top end you need to roll off and you like distortion, but its a pretty advanced area to talk about.

So for the 99.9% that couldn't care less about adding 50% more cash into their system for the tuning ability of a full active system (and hopefully have the ear or people to assist in tuning it) its not anything to worry about.

Richie - we're all nerds, we just admit it.

I'll PM you about some topics in a little bit here.